vix.ing · top · new · best · stats · spec

Asymptotic Inference for Exchangeable Gibbs Partitions

2025/06/26 by Takuya Koriyama, Koriyama, Takuya
Mathematics · Computer Science · #Markov Chains and Monte Carlo Methods #Bayesian Methods and Mixture Models

paper · pdf · doi:10.48550/arxiv.2506.21527

Abstract

We study the asymptotic properties of parameter estimation and predictive inference under the exchangeable Gibbs partition, characterized by a discount parameter α∈(0,1) and a triangular array vn,k satisfying a backward recursion. Assuming that vn,k admits a mixture representation over the Ewens--Pitman family (α, θ), with θ integrated by an unknown mixing distribution, we show that the (quasi) maximum likelihood estimator αn (QMLE) for α is asymptotically mixed normal. This generalizes earlier results for the Ewens--Pitman model to a more general class. We further study the predictive task of estimating the probability simplex pn, which governs the allocation of the (n+1)-th item, conditional on the current partition of [n]. Based on the asymptotics of the QMLE αn, we construct an estimator pn and derive the limit distributions of the f-divergence Df(pn||pn) for general convex functions f, including explicit results for the TV distance and KL divergence. These results lead to asymptotically valid confidence intervals for both parameter estimation and prediction.

Citations

Related